Automated fact-checking, using machine learning to verify claims, has grown vital as misinformation spreads beyond human fact-checking capacity. Large Language Models (LLMs) like GPT-4 are increasingly trusted to write academic papers, lawsuits, and news articles and to verify information, emphasizing their role in discerning truth from falsehood and the importance of being able to verify their outputs. Understanding the capacities and limitations of LLMs in fact-checking tasks is therefore essential for ensuring the health of our information ecosystem. Here, we evaluate the use of LLM agents in fact-checking by having them phrase queries, retrieve contextual data, and make decisions. Importantly, in our framework, agents explain their reasoning and cite the relevant sources from the retrieved context. Our results show the enhanced prowess of LLMs when equipped with contextual information. GPT-4 outperforms GPT-3, but accuracy varies based on query language and claim veracity. While LLMs show promise in fact-checking, caution is essential due to inconsistent accuracy. Our investigation calls for further research, fostering a deeper comprehension of when agents succeed and when they fail.

In this critical survey, we analyze typical claims on the relationship between explainable AI (XAI) and fairness to disentangle the multidimensional relationship between these two concepts. Based on a systematic literature review and a subsequent qualitative content analysis, we identify seven archetypal claims from 175 papers on the alleged fairness benefits of XAI. We present crucial caveats with respect to these claims and provide an entry point for future discussions around the potentials and limitations of XAI for specific fairness desiderata. Importantly, we notice that claims are often (i) vague and simplistic, (ii) lacking normative grounding, or (iii) poorly aligned with the actual capabilities of XAI. We encourage to conceive XAI not as an ethical panacea but as one of many tools to approach the multidimensional, sociotechnical challenge of algorithmic fairness. Moreover, when making a claim about XAI and fairness, we emphasize the need to be more specific about what kind of XAI method is used and which fairness desideratum it refers to, how exactly it enables fairness, and who is the stakeholder that benefits from XAI.

The proliferation of large AI models trained on uncurated, often sensitive web-scraped data has raised significant privacy concerns. One of the concerns is that adversaries can extract information about the training data using privacy attacks. Unfortunately, the task of removing specific information from the models without sacrificing performance is not straightforward and has proven to be challenging. We propose a rather easy yet effective defense based on backdoor attacks to remove private information such as names and faces of individuals from vision-language models by fine-tuning them for only a few minutes instead of re-training them from scratch. Specifically, through strategic insertion of backdoors into text encoders, we align the embeddings of sensitive phrases with those of neutral terms-"a person" instead of the person's actual name. For image encoders, we map embeddings of individuals to be removed from the model to a universal, anonymous embedding. Our empirical results demonstrate the effectiveness of our backdoor-based defense on CLIP by assessing its performance using a specialized privacy attack for zero-shot classifiers. Our approach provides not only a new "dual-use" perspective on backdoor attacks, but also presents a promising avenue to enhance the privacy of individuals within models trained on uncurated web-scraped data.

The presenter pasted in what he called "about 16 pages' worth of tax code" These seven sentences about Alice, Bob, and Charlie come word-for-word from a handcrafted data set we developed at Johns Hopkins University and published in 2020 for training and measuring AI models for reasoning over statutory language. Every word, punctuation mark, and Maryland; Nils number in the taxpayer facts comes exactly from Holzenberger is an our tax_case_9 -- even the percent sign at the start associate professor in of the line. This work has been supported by the U.S. National Science Foundation under grant No. 2204926. The entire livestream is available at OpenAI, "GPT-4 Developer The tax law example starts at minute 19:11. Go to the directory "Cases" to find the file tax_case_9.pl. Tax_case_9.pl is written in the programming language Prolog. Federal content, please visit www.taxnotes.com. Where did the "about 16 pages' worth of tax out the TCJA standard deduction increase at code" come from? Again, from our 2020 data set. SARA has two deduction for 2018 was $24,000. From minute 20:07 to 20:40 of the livestream, handcrafted cases in SARA; tax_case_9 is one of we see some of the tax sections pasted into GPT-4. The statutes consist of nine sections of the These are SARA's heavily edited version of the IRC, For example, at and remove ambiguity. If you put all the SARA 20:23, we see part of section 63(c) with the statutes into a single file it will be about 16 pages paragraphs jumping from (3) to (5); in SARA, we long (depending on the font). At 20:26, we see part of section One of our edits was paring section 1 down to 63(c)(6) with only subparagraphs (A), (B), and (D); only sections 1(a) through (d), which contain the in SARA, we edited out (C). At 20:40, we see parts Clinton-era tax rates. We cut section 1(j), which of section 3306(b) with the paragraphs jumping contains the reduced Tax Cuts and Jobs Act rates from (2) to (7); in SARA, we edited out paragraphs for 2018-2025. This editing explains why GPT-4 (3) through (6). At 20:39 we see sections 3301 and got the wrong answer on the livestream for Alice 3306 regarding the federal unemployment tax; and Bob's 2018 taxes. We did not, however, edit while these two sections are irrelevant to Alice and Bob's tax liability in tax_case_9, they are two The author Holzenberger did all the handcrafting and hand editing. Federal content, please visit www.taxnotes.com. You can We empirically verified that using the SARA download our data set and compare it with the version of the IRC causes GPT-4 to get the wrong livestream's recording on YouTube. First, we The presenter then gives directions to GPT-4: pasted into GPT-4 all nine SARA statutes, plus our "Now calculate their total liability." GPT-4 gives facts about Alice, Bob, and Charlie. Then we detailed step-by-step calculations and concludes used the same "Now calculate their total liability" that "Alice and Bob's total tax liability for 2018 is command.

Although ubiquitous in modern vehicles, Controller Area Networks (CANs) lack basic security properties and are easily exploitable. A rapidly growing field of CAN security research has emerged that seeks to detect intrusions on CANs. Producing vehicular CAN data with a variety of intrusions is out of reach for most researchers as it requires expensive assets and expertise. To assist researchers, we present the first comprehensive guide to the existing open CAN intrusion datasets, including a quality analysis of each dataset and an enumeration of each's benefits, drawbacks, and suggested use case. Current public CAN IDS datasets are limited to real fabrication (simple message injection) attacks and simulated attacks often in synthetic data, which lack fidelity. In general, the physical effects of attacks on the vehicle are not verified in the available datasets. Only one dataset provides signal-translated data but not a corresponding raw binary version. Overall, the available data pigeon-holes CAN IDS works into testing on limited, often inappropriate data (usually with attacks that are too easily detectable to truly test the method), and this lack data has stymied comparability and reproducibility of results. As our primary contribution, we present the ROAD (Real ORNL Automotive Dynamometer) CAN Intrusion Dataset, consisting of over 3.5 hours of one vehicle's CAN data. ROAD contains ambient data recorded during a diverse set of activities, and attacks of increasing stealth with multiple variants and instances of real fuzzing, fabrication, and unique advanced attacks, as well as simulated masquerade attacks. To facilitate benchmarking CAN IDS methods that require signal-translated inputs, we also provide the signal time series format for many of the CAN captures. Our contributions aim to facilitate appropriate benchmarking and needed comparability in the CAN IDS field.

A domain adaptation technique has been proposed in this paper to identify the emotions in generic images containing facial & non-facial objects and non-human components. It addresses the challenge of the insufficient availability of pre-trained models and well-annotated datasets for image emotion recognition (IER). It starts with proposing a facial emotion recognition (FER) system and then moves on to adapting it for image emotion recognition. First, a deep-learning-based FER system has been proposed that classifies a given facial image into discrete emotion classes. Further, an image recognition system has been proposed that adapts the proposed FER system to recognize the emotions portrayed by images using domain adaptation. It classifies the generic images into 'happy,' 'sad,' 'hate,' and 'anger' classes. A novel interpretability approach, Divide and Conquer based Shap (DnCShap), has also been proposed to interpret the highly relevant visual features for emotion recognition. The proposed system's architecture has been decided through ablation studies, and the experiments are conducted on four FER and four IER datasets. The proposed IER system has shown an emotion classification accuracy of 59.61% for the IAPSa dataset, 57.83% for the ArtPhoto dataset, 67.93% for the FI dataset, and 55.13% for the EMOTIC dataset. The important visual features leading to a particular emotion class have been identified, and the embedding plots for various emotion classes have been analyzed to explain the proposed system's predictions.

Two companies based in Texas have been linked to a spate of robocalls that used artificial intelligence to mimic President Joe Biden. The audio deepfake was used to urge New Hampshire voters not to participate in the state's presidential primary. New Hampshire Attorney General John Formella said as many as 25,000 of the calls were made to residents of the state in January. Formella says an investigation has linked the source of the robocalls to Texan companies Life Corporation and Lingo Telecom. No charges have yet been filed against either company or Life Corporation's owner, a person named Walter Monk.

Meta is introducing a tool to identify AI-generated images shared on its platforms amid a global rise in synthetic content spreading misinformation. Due to several of systems on the web, the Mark Zuckerberg-owned company is aiming to expand labels to others like Google, OpenAI, Microsoft, and Adobe. Meta said it will fully roll out the labeling feature in the coming months and plans to add a feature that lets users flag AI-generated content. However, the US presidential race is in full swing, leaving some to wonder if the labels will be out in time to stop fake content from spreading. The move comes after Meta's Oversight Board urged the company to take steps to label manipulated audio and video that could mislead users. 'The Board's recommendations go further in that it advised the company to expand the Manipulated Media policy to include audio, clearly state the harms it seeks to reduce, and begin labeling these types of posts more broadly than what was announced,' an Oversight Board spokesperson Dan Chaison told Dailymail.com.

Science fiction, when revisited years later, sometimes doesn't come across as all that fictional. Speculative novels have an impressive track record at prophesying what innovations are to come, and how they might upend the world: H. G. Wells wrote about an atomic bomb decades before World War II, and Ray Bradbury's 1953 novel, Fahrenheit 451, features devices we'd describe today as Bluetooth earbuds. Perhaps no writer has been more clairvoyant about our current technological age than Neal Stephenson. His novels coined the term metaverse, laid the conceptual groundwork for cryptocurrency, and imagined a geoengineered planet. A core element of one of his early novels, The Diamond Age: Or, a Young Lady's Illustrated Primer, is a magical book that acts as a personal tutor and mentor for a young girl, adapting to her learning style--in essence, it is a personalized and ultra-advanced chatbot.

Meta plans to ramp up its labeling of AI-generated images across Facebook, Instagram and Threads to help make it clear that the visuals are artificial. It's part of a broader push to tamp down misinformation and disinformation, which is particularly significant as we wrangle with the ramifications of generative AI (GAI) in a major election year in the US and other countries. According to Meta's president of global affairs, Nick Clegg, the company has been working with partners from across the industry to develop standards that include signifiers that an image, video or audio clip has been generated using AI. "Being able to detect these signals will make it possible for us to label AI-generated images that users post to Facebook, Instagram and Threads," Clegg wrote in a Meta Newsroom post. "We're building this capability now, and in the coming months we'll start applying labels in all languages supported by each app."